Add doctests to Perceiver examples (#19129)

* Fix bug in example and add to tests

* Fix failing tests

* Check the size of logits

* Code style

* Try again...

* Add expected loss for PerceiverForMaskedLM doctest

Co-authored-by: Steven Anton <antonstv@amazon.com>
Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>

src/transformers/models/perceiver/modeling_perceiver.py

@@ -801,6 +801,8 @@ def forward(
         >>> with torch.no_grad():
         ...     outputs = model(inputs=inputs)
         >>> logits = outputs.logits
+        >>> list(logits.shape)
+        [1, 2]
 
         >>> # to train, one can train the model using standard cross-entropy:
         >>> criterion = torch.nn.CrossEntropyLoss()
@@ -810,6 +812,7 @@ def forward(
 
         >>> # EXAMPLE 2: using the Perceiver to classify images
         >>> # - we define an ImagePreprocessor, which can be used to embed images
+        >>> config = PerceiverConfig(image_size=224)
         >>> preprocessor = PerceiverImagePreprocessor(
         ...     config,
         ...     prep_type="conv1x1",
@@ -844,6 +847,8 @@ def forward(
         >>> with torch.no_grad():
         ...     outputs = model(inputs=inputs)
         >>> logits = outputs.logits
+        >>> list(logits.shape)
+        [1, 2]
 
         >>> # to train, one can train the model using standard cross-entropy:
         >>> criterion = torch.nn.CrossEntropyLoss()
@@ -1017,7 +1022,12 @@ def forward(
 
         >>> outputs = model(**inputs, labels=labels)
         >>> loss = outputs.loss
+        >>> round(loss.item(), 2)
+        19.87
+
         >>> logits = outputs.logits
+        >>> list(logits.shape)
+        [1, 2048, 262]
 
         >>> # inference
         >>> text = "This is an incomplete sentence where some words are missing."
@@ -1030,6 +1040,8 @@ def forward(
         >>> with torch.no_grad():
         ...     outputs = model(**encoding)
         >>> logits = outputs.logits
+        >>> list(logits.shape)
+        [1, 2048, 262]
 
         >>> masked_tokens_predictions = logits[0, 52:61].argmax(dim=-1).tolist()
         >>> tokenizer.decode(masked_tokens_predictions)
@@ -1128,6 +1140,8 @@ def forward(
         >>> inputs = tokenizer(text, return_tensors="pt").input_ids
         >>> outputs = model(inputs=inputs)
         >>> logits = outputs.logits
+        >>> list(logits.shape)
+        [1, 2]
         ```"""
         if inputs is not None and input_ids is not None:
             raise ValueError("You cannot use both `inputs` and `input_ids`")
@@ -1265,9 +1279,13 @@ def forward(
         >>> inputs = feature_extractor(images=image, return_tensors="pt").pixel_values
         >>> outputs = model(inputs=inputs)
         >>> logits = outputs.logits
+        >>> list(logits.shape)
+        [1, 1000]
+
         >>> # model predicts one of the 1000 ImageNet classes
         >>> predicted_class_idx = logits.argmax(-1).item()
         >>> print("Predicted class:", model.config.id2label[predicted_class_idx])
+        Predicted class: tabby, tabby cat
         ```"""
         if inputs is not None and pixel_values is not None:
             raise ValueError("You cannot use both `inputs` and `pixel_values`")
@@ -1402,9 +1420,13 @@ def forward(
         >>> inputs = feature_extractor(images=image, return_tensors="pt").pixel_values
         >>> outputs = model(inputs=inputs)
         >>> logits = outputs.logits
+        >>> list(logits.shape)
+        [1, 1000]
+
         >>> # model predicts one of the 1000 ImageNet classes
         >>> predicted_class_idx = logits.argmax(-1).item()
         >>> print("Predicted class:", model.config.id2label[predicted_class_idx])
+        Predicted class: tabby, tabby cat
         ```"""
         if inputs is not None and pixel_values is not None:
             raise ValueError("You cannot use both `inputs` and `pixel_values`")
@@ -1539,9 +1561,13 @@ def forward(
         >>> inputs = feature_extractor(images=image, return_tensors="pt").pixel_values
         >>> outputs = model(inputs=inputs)
         >>> logits = outputs.logits
+        >>> list(logits.shape)
+        [1, 1000]
+
         >>> # model predicts one of the 1000 ImageNet classes
         >>> predicted_class_idx = logits.argmax(-1).item()
         >>> print("Predicted class:", model.config.id2label[predicted_class_idx])
+        Predicted class: tabby, tabby cat
         ```"""
         if inputs is not None and pixel_values is not None:
             raise ValueError("You cannot use both `inputs` and `pixel_values`")
@@ -1689,6 +1715,8 @@ def forward(
         >>> patches = torch.randn(1, 2, 27, 368, 496)
         >>> outputs = model(inputs=patches)
         >>> logits = outputs.logits
+        >>> list(logits.shape)
+        [1, 368, 496, 2]
         ```"""
         return_dict = return_dict if return_dict is not None else self.config.use_return_dict
 
@@ -1915,6 +1943,14 @@ def forward(
 
         >>> outputs = model(inputs=inputs, subsampled_output_points=subsampling)
         >>> logits = outputs.logits
+        >>> list(logits["audio"].shape)
+        [1, 240]
+
+        >>> list(logits["image"].shape)
+        [1, 6272, 3]
+
+        >>> list(logits["label"].shape)
+        [1, 700]
         ```"""
         return_dict = return_dict if return_dict is not None else self.config.use_return_dict
 
@@ -2925,7 +2961,6 @@ def __init__(self, config: PerceiverConfig, in_channels: int, postproc_type: str
         self.classifier = nn.Linear(in_channels, config.samples_per_patch)
 
     def forward(self, inputs: torch.Tensor, pos: Optional[torch.Tensor] = None, modality_sizes=None) -> torch.Tensor:
-
         logits = self.classifier(inputs)
         return torch.reshape(logits, [inputs.shape[0], -1])
 

utils/documentation_tests.txt

@@ -58,6 +58,7 @@ src/transformers/models/opt/modeling_opt.py
 src/transformers/models/opt/modeling_tf_opt.py
 src/transformers/models/owlvit/modeling_owlvit.py
 src/transformers/models/pegasus/modeling_pegasus.py
+src/transformers/models/perceiver/modeling_perceiver.py
 src/transformers/models/plbart/modeling_plbart.py
 src/transformers/models/poolformer/modeling_poolformer.py
 src/transformers/models/reformer/modeling_reformer.py